The FDA reported a new streamlined pathway for diagnostic testing as part of a serial testing program using pooled samples, a pathway that should enable the further reopening of the economy. However, the FDA’s Tim Stenzel said April 21 that this new pathway relied on accumulated data for molecular testing, and that the agency lacks sufficient data to provide a similar mechanism for this use of antigen tests.

The FDA posted a novel pathway for serial test pooling based on molecular tests performed in labs certified to handle high-complexity tests, and Stenzel, director of the Office of In Vitro Diagnostics and Radiological Health (OIR) at the FDA, noted that this new pathway is available for tests that are already authorized for use with pooled anterior nasal specimens for screening. This assumes that individuals will be part of a pooled test once a week, although the conditions of this pathway vary somewhat, depending on the number of samples to be included in a pooled sample.

During the April 21 COVID testing town hall, Stenzel said the April 20 pooling framework was enabled in large part by the volume of emergency use authorizations (EUAs) for pooling and for screening tests. All the tests authorized for pooling have been high-sensitivity molecular assays for central lab processing, and Stenzel said the agency has learned a lot from those EUAs, which allowed the agency to offer the new streamlined pathway. The agency has not seen data indicative of performance issues for pooling three samples or more, if the test in question is a high-sensitivity molecular test for analysis in central labs.

When it comes to antigen tests, however, Stenzel said a similar policy is inhibited in part by the absence of an authorized antigen test for pooling. “It would be nice to see an application where the sensitivity is great enough that we can pool,” he said, but the lack of data on the performance characteristics for antigen tests for this use is an insuperable hurdle, at least for the time being.

Any molecular test thus authorized must be accompanied by a pooling protocol that provides instructions for follow-up for a positive result and for invalid pooling. This includes deconvolution for retest of individual samples. One of the challenges associated with swab pooling is that immediate deconvolution requires that two swabs be collected from each participant. Stenzel said some developers have expressed an interest in a testing regime that relies on swab pooling, and that there is a lower risk of false negatives when using swab pooling.

High-volume, OTC tests gaining traction

The agency’s emphasis on high-volume tests is a matter of record, and the FDA recently authorized an amplitude solution with the Taqpath high-throughput combination kit from Thermo Fisher Scientific Inc. This test can be used to process as many as 8,000 reactions over a period of 24 hours. The agency also expanded the EUA for the Binax Now self-test by Abbott Laboratories for use without a prescription, which the company will sell in packs of two at drug stores. The FDA also reported a new EUA for the Synergy Dx direct-to-consumer sample kit by Synergy Diagnostic Lab Inc., of Davie, Fla. This sample collection kit must be sent to a lab for processing, but the company also received an EUA for the underlying test, a PCR test authorized for use in both symptomatic and asymptomatic individuals.

Toby Lowe, associate director of OIR, said the FDA is not yet persuaded about the use of contrived samples for initial authorization of a multi-analyte test. The agency would, in this instance, expect to see a prospective study that uses an authorized PCR test as a comparator for the SARS-CoV-2 virus and an FDA-cleared test as a comparator for the other claimed analytes.

Lowe said that the agency is generally amenable to the use of data from sites outside the U.S. (OUS) for test validation, but there are exceptions. One of these is tests used at the point of care, and the other is a test or sample that requires home validation studies. This is because the agency is aware of differences in how the tests and test kits perform across nations, but she indicated that a U.S. data set could be supplemented by data from other nations, such as Canada. The problem for POC tests is that the FDA cannot be certain about the standard practice at these OUS sites.

CDC reports 30k new sequences

The CDC has been keeping a running tally on the number of published sequences for the pandemic, and the latest report shows that more than 30,000 sequences were reported for the week of April 17. That number includes more than 23,000 sequences handled by the agency, and is double the number sequenced only two weeks earlier.

Those numbers do not include sequences turned out by public health labs or those generated in fulfillment of academic research contracts. As matters stand, a total of roughly 450,000 viral sequences are publicly available to test developers from U.S. sources, with the state of California providing more than 41,000 of those sequences since the start of the pandemic.

The problem of cases in which a variant of the virus has affected vaccinated individuals is gaining traction in U.S. federal government circles, with the CDC noting that state and local officials can enter data from these sequences in its REDcap database. Eventually, CDC will use the National Notifiable Disease Surveillance System (NNDSS) to identify these breakthrough cases, once states are able to report vaccination history to the NNDSS.